In recent years, mobile devices such as cell-phones (and in particular smart-phones), tablets and laptops have become ubiquitous. Most of these devices include one or two compact cameras: a main rear-facing camera (i.e. a camera on the back side of the device, facing away from the user and often used for casual photography) and a secondary front-facing camera (i.e. a camera located on the front side of the device and often used for video conferencing).
Although relatively compact in nature, the design of most of these cameras is very similar to the traditional structure of a digital still camera, i.e. they comprise an optical component (or a train of several optical elements and a main aperture) placed on top of an image sensor. The optical component (also referred to as “optics”) refracts the incoming light rays and bends them to create an image of a scene on the sensor. The dimensions of these cameras are largely determined by the size of the sensor and by the height of the optics. These are usually tied together through the focal length (“f”) of the lens and its field of view (FOV)—a lens that has to image a certain FOV on a sensor of a certain size has a specific focal length. Keeping the FOV constant, the larger the sensor dimensions (e.g. in a X-Y plane) the larger the focal length and the optics height.
In addition to the optics and sensor, modern cameras usually further include mechanical motion (actuation) mechanism for two main purposes: focusing of the image on the sensor and optical image stabilization (OIS). For focusing, in more advanced cameras, the position of the lens module (or at least one lens element in the lens module) can be changed by means of an actuator and the focus distance can be changed in accordance with the captured object or scene. In these cameras it is possible to capture objects from a very short distance (e.g., 10 cm) to infinity. The trend in digital still cameras is to increase the zooming capabilities (e.g. to 5×, 10× or more) and, in cell-phone (and particularly smart-phone) cameras, to decrease the pixel size and increase the pixel count. These trends result in greater sensitivity to hand-shake or in a need for longer exposure time. An OIS mechanism is required to answer the needs in these trends.
In OIS-enabled cameras, the lens or camera module can change its lateral position or tilt angle in a fast manner to cancel the handshake during the image capture. Handshakes move the camera module in 6 degrees of freedom, namely linear movements in three degrees of freedom (X, Y & Z), pitch (tilt around the X axis), yaw (tilt around the Y axis) and roll (tilt around the Z axis). FIG. 1 shows an exemplary classical four rod-springs (102a-d) OIS structure in a single-aperture camera module 100. The four rod-springs are rigidly connected to an upper frame 104 that usually accommodates an AF actuator (not shown) that moves the lens module 106. This structure allows desired modes of movement in the X-Y plane (translation), FIG. 1a, but also allows a mode of unwanted rotation (also referred to as “θ-rotation” or “torsion”) around the Z axis, FIG. 1b. The latter may be due to a combination of several causes such as asymmetric forces applied by the coils or by a user's (or phone) movements, imperfections of the rod-springs and the high rotational compliance of the four spring rod spring+frame structure.
In the case of a centered single-aperture camera module, this rotation does not affect the image quality severely, since the lens is axisymmetric. However, this does affect OIS in a dual-camera module, FIGS. 2a and 2b. FIG. 2a shows a rotation mode around an axis 202 roughly centered between two camera modules 204 and 206 of a dual-aperture camera 200. Because of the location of rotation axis 202, the rotation may cause significant deterioration in the image quality. The rotation causes each lens to shift away in undesired directions (shown by arrows in FIG. 2b), without having any ability to predict when and if this may happen. The result is motion blur of the image and a shift of the two lenses in opposite Y directions caused by the unwanted rotation that results in decenter between images received by each camera module, and therefore potentially in a catastrophic influence on fusion algorithm results.
Yet another problem may occur in a folded optics zoom dual-aperture camera, such as a camera 300 shown in FIG. 3. Such a camera is described for example in detail in co-owned international patent application PCT/IB2016/052179 which is incorporated herein by reference in its entirety. Camera 300 comprises a folded optics camera module 302 and an upright (non-folded) camera module 304. Among other components, folded optics camera module 302 comprises a lens actuation sub-assembly for moving a lens module 306 (and a lens therein, which is referred to henceforth as “folded lens”) in the X-Y plane. The lens actuation sub-assembly includes a hanging structure with four flexible hanging members (i.e. the “rod-springs” referred to above) 308 a-d that hang lens module 306 over a base 310. In some embodiments, hanging members 308a-d may be in the form of four wires and may be referred to as “wire springs” or “poles”. The hanging structure allows in-plane motion as known in the art and described exemplarily in co-owned U.S. patent application Ser. No. 14/373,490. Exemplarily, a first movement direction 312 of the lens is used to achieve Auto-Focus (AF) and a second movement direction 314 is used to achieve OIS. A third movement, an unwanted rotation 316 of the lens about an axis parallel to the Z axis as described above actually causes an unwanted effect of dynamic tilt of the lens (the lens' optical axis may not be perpendicular to the sensor's surface due to that rotation) and may result in images that are usually sharp on one side and blurry on the other side.
The physical quantities that reflect the tendency of any structure to dynamically behave one way or another are the natural frequency values that characterize each mode of behavior. This is of course also relevant for the hanging structure described above. FIGS. 4(a)-(c) show the simulated behavior of a standard rigid plate supported by four round rod-spring poles. The rigid plate may represent any optical element (such as, for example, a lens). The rod-spring poles have the same rigidity to movement in any direction in the X-Y plane (which is perpendicular to the pole's neutral axis). The figures show the compliance of the structure expressed in terms of a natural frequency ratio for each different movement mode; FIG. 4a refers to X-translation, FIG. 4b refers to Y-translation and FIG. 4c refers to rotation around the Z axis. The arrows show schematically the different movements. The reference bar indicates deformation scale in millimeters. The normalized (relative to the first frequency which in this exemplary case is of 33.6 Hz) natural frequencies for X and Y translations are of the same order (specifically 1 in (a) and 1.1 in (b)), whilst the natural frequency for rotation (c) has a relative value of 1.8, which is also of the same order of the X and Y translations. Thus the ratio between natural frequencies for torsion (rotation around Z) and for X or Y translation is about 1.8. In general, known ratios are no larger than 2. This means that the chance that the torsion mode will arise is almost the same as the chance that the X and Y translation modes will arise. This may cause problems in dual-aperture and/or folded zoom cameras (where it will be expressed as dynamic tilt) as described above.
In view of the above, it would be very difficult to get the desired movement of the lens without an active control loop (having such a control loop is one possible way to overcome the described problems). The unwanted torsion may be reduced significantly by means of electrical control over the force applied by the coils (i.e. by using several coils and controlling them so the resultant torque acts to limit the rotation of the lens within specified acceptable limits). However, the addition of an active control loop to avoid tilt complicates the design and adds to cost. It would be therefore advantageous to have lens actuation sub-assemblies for OIS without an active control loop for rotation/tilt.